's column on
Metcalfe's Law in Reverse
Who is Metcalfe?
Emery Jeffreys from
For the most part I get quite a bit out of your Alertbox items. It often prompts me to change my mind on net issues. It always makes me think about the issue.
However, on the current issue involving Metcalfe's law, I was a little lost. Just which Metcalfe to whom were you referring? Robert Metcalfe, the inventor of the ethernet card? Or was it a math pro who has a law named after him?
Generally in writings distributed outside of academia to the public, you should identify on first reference, the person by first name. Referring to a person only by their last names is reserved for people who have preeminence -- like Einstein, Churchill, Roosevelt or Hitler. I doubt if Metcalfe fits that category.
I'm also not sure if I agree with Cisco's assessment that the net is worth $300 billion, not can I agree with your observations in diminishing the financial value of the web because of the actions of its participants.
Your argument is weighing the future value of the net against the present value. And that must take into account for economic inflation to be practical.
At one point you estimate that the future value of the Web would be reduced from $1 Trillion to $100 Billion - for a loss to society of $900 Billion.
Society has yet to understand and realize the value of the net. Without that realization, it is impossible for them to understand the loss.
Metcalfe's Law is indeed named after
Dr. Robert M. Metcalfe
, the inventor of Ethernet and columnist for
. When I was writing the original Alertbox, I actually wanted to link Metcalfe's name to his home page, but I simply couldn't find it. Still can't. Instead, here is a link to his profile in
The Industry Standard
's "Who's Who of the Internet Economy" (always a good place to find bios of people in this field).
I believe that the term "Metcalfe's Law" was introduced in a 1993 article by George Gilder in
magazine entitled Metcalfe's Law and Legacy.
It is hard to say what the Web and Internet are worth now, but I find the Cisco study reasonably credible as a rough estimate.
Here is a completely different way of
estimating the value of the Web
The Web currently has about 180 million users
The average user spends 11 hours per month online
Thus, the Web consumes 24 billion person-hours per year
Assume that a user's time is worth $12.50 per hour
The value of the time spent using the Web is $300 Billion
This does not count intranet usage, which probably sums to $200 Billion across companies (and much of it
wasted due to poor intranet usability
Many peoples' time is surely worth more than $12.50 per hour, since Web usage tends to be concentrated among the affluent (and since you need to look at the
cost of an employee, including benefits and overhead, and not simply their salary). On the other hand, it is doubtful whether users currently get sufficient return in terms of results for the time they spend on the Web. But since it is completely voluntary to use the Web, users must
that they get value for their time. So even though there is no single accounting statement that says "$300 Billion" on the bottom line, the collective actions of the world's users indicate that they value the Web at about that much.
Who Keeps the Value?
Arnt Gulbrandsen from
This Alertbox assumes that the value of the web accrues to web site owners. This is not correct - the web is valuable to clients too.
Metcalfe's law says something about the total worth of a network; you assume that it says something about the value of providing content on a network.
You are right that not all the value goes to the producers of a service. In the long term, people will only use something if they get
value from it than the cost. So the overall value splits in some way into two parts: the ability of the provider to "monetize" its traffic and the residue surplus to the users.
Take as an example the question of placing a phone call to your mother. Let's assume that the phone call will cost $1. That's the phone company's immediate earnings from providing the service. It is worth more than $1 to you to talk to your mother? If yes, you place the call. If no, you don't. To the extent that you value the conversation at more than $1, the surplus is a benefit you get to keep. Also, if your mother thinks that she gets positive value out of talking to you (one would hope so), then she is allowed to keep that surplus value. Of course, both you and your mother have to pay the phone company and the telephone manufacturer for the monthly cost of getting service and the purchase cost of the telephone sets used for the call. So parts of your immediate surplus still goes back to the producers.
Even so, much of the value should accrue to the producer if they manage their service correctly. For example, phone calls may be priced very low these days due to competition and technological advances. So the person placing a truly business-critical call will profit immensely and get to keep a huge surplus value beyond the cost of the call. But at the same time the lower cost of telephony leads to
teenagers gabbing on cell phones
at all times of day, creating new business opportunities.
Correct, we have to speak about "value to society" when discussing the Internet because
of the value goes to the service providers, but not all. But once we get micropayments, the providers should be able to keep a good deal of the value they create.
Proprietary Networks Can Have Value
Brian Hayashi writes:
True, the total utility to the user diminishes if the network fails to take into account the full diversity of the Net. However:
The money that is lost is really the money that is invested in competitive business models. You would actually prefer your competitors to lose money because of your site. (I've probably read too much Guy Kawasaki.)
This assumes all utility is publisher-to-individual. Business-to-business networks, such as Mastercard, compete based on proprietary utility/features of the network.
There are natural constraints to how much content can be leveraged and the ability to switch smoothly from one business model to the next. Yahoo is an example of this. Their information architects have been overloaded and despite the wonders of technology, it would be difficult for them to switch to a Google model, for example.
We already have situations where sites try to leverage as much of the Web as they can. The result is a number of sites that look and act pretty much the same. Business valuation concepts are dependent on the ability of a business to differentiate itself from its competition in a proprietary kind of way.
Of course, nothing is really ever black and white. We want to appear open while simultaneously trying to shut the door behind us. Microsoft led this tactic, and most recently, AOL emulated this by demanding access to AT&T while claiming something entirely different on ICQ.
I really like your site and the lessons it has on it. We don't always follow conventions because our target market has specific requirements and we don't feel the need to let other Internet companies know what we're up to.
love the idea of a business strategy that causes your competitors to lose money, even if it doesn't do anything else.
I also agree that proprietary networks can have
value; even a good deal of value as in the case of Mastercard. The question is simply whether they would have more value if they were open and bigger. A credit card network obviously needs high security which we don't have on the Web today - but we
have it if (a) encryption was freed and (b) software vendors and the operating system vendor would make pervasive security a higher priority.
So, assuming a secure Internet, could Mastercard get more value from being an open service than from running their own proprietary network? Yes, because now it becomes possible to use their payment system for many more kinds of transactions, maybe even things like getting a Coke from a soft drink machine.
As another example, compare EDI (electronic data interchange) with XML. EDI is not strictly proprietary since the specifications are based on official standards, but in practice it is only used for small, closed networks. Because of the closed nature of EDI, it is almost never used: big auto manufacturers do use EDI to manage payments and order flows to their parts vendors, but that's the exceptional case. Almost all transactions involve somebody typing the data into a computer, an invoice being printed and mailed, and somebody else typing the same data into another computer on the other end to cause a a check to be printed out and mailed back (after which a third person types the data into a computer once again). XML is hoped to be a more widely used solution that can automate a much larger set of transactions than was ever possible with EDI. Thus, the value of XML will likely be enormously larger than the value of EDI.
Freedom of the Press
Emery Jeffreys from
The latest alertbox on Metcalfe's law was on the mark about the ways that many content sites are trying to undermine universal connectivity. It really amounts to posturing and position taking.
Ultimately, they will realize the error of their ways when they discover it is costing them customers.
Average web users are too clueless about the issues to know that they should be raising hell about it. I hate to say it like that, but it is true.
I think the issues regarding the media are side issue of their own and really constitute a
free press issue
-- one that they may not be equipped to deal with.
When I was managing editor of The Daytona Beach News-Journal web site, Daytona International Speedway refused to issue press credentials to me because I was going to generate stories for a web site, not the print edition of the paper.
(My qualifications to receive credential were not an issue. Up until I left the newspaper business 16 months ago I had more than 25 years experience -- including working as an editor at United Press International)
In past years at the newspaper, we routinely had results of the Daytona 500 race on our web site before NASCAR had the results on its own web site.
My senior editor (from the old school) quite simply blustered them. They were smart enough not to call his bluff.
Another onerous example is the NCAA, the agency that regulates college sports. For a time, they would not issue press credentials to newspapers that published handicapping services of sports teams.
That's a real two-faced example because the NCAA from time-to-time has used
The Dunkel Index
to seed (choose) tournament rankings in Division II.
Online news sources are going to have hard time making a case here because all they do is shovel their existing content on to the web. Web users look to the other sites because it provides something not available in newspaper web sites.
I agree that the Web raises new issues about the freedom of the press. Or, rather, they are the
old issues, but in new clothing
, making them hard to see for some people who are still focused on the old media world.
We need to take care to preserve the values that served us well in the past. Paradoxically, I think that newspapers could be well-positioned to become some of the most trusted and valuable services on the Web if only they would manage the transition from print to online correctly and aggressively. But I don't see many signs of this happening. Most publishers still don't agree with my statement that
print will be dead in 10 years
, so they are not doing what it takes.